In the early seventies I first became involved with computer control of industrial processes during my student days at Dunlop Ltd. in Birmingham . I spent a summer break at the factory working in the first computer controlled rubber compounding plant in the UK. The plant was run by a mini computer in its own air conditioned room (about 75 m2) which had less processing power than a current games machine. Starting up involved loading the bootstrap program into the machine, by hand on digital switches, this provided the code to operate the paper tape reader which could then be used to read in the rest of the operating code. The programs and patches were all stored on punched paper tape a very unstable medium in an industrial environment, and the tapes had to be loaded in the correct order, miss out just one tape and you started again..
A few years later I had moved to the South west and had my first experience of Proggrammable Logic Controllers (PLC for short), these were programmed using dedicated programming panels and billed as relay replacements, though we are still using relays today over thirty years later.
We have been involved with the IOT for several years gaining valuable experience whilst designing a web server framework to:
1/ Collect data from external GPRS and Ethernet enabled modems using encoded transmissions on a timed basis allowing data to be read every 30 minutes as standard. We can actually get down to around every two seconds when imiting the quantity of data on a single communications channel.
2/ To store the data for long term analysis and display.
3/ To provide customer accounts that allow access to the data using graphical displays allowing year on year comparison of energy consumption.
This framework is used by several web sites to allow their customers to access their individual energy data. Customers include Solar Energy companies, property landlords, Super Market chains and many others.
I became involved with micro processors about this time building one of the Science of Cambridge kits, an 8 bit micro-processor with just 256 bytes of ram. Later I progressed to designing my own computer using a Signetics 2650 micro processor, an unsual beast in that it could be driven by a DC clock, I built it with a simple switch to drive the clock line and could watch each step of the processor via LED's on the data and address lines, you learn a lot from this type of experience. I moved on from hand coding machine code programs to assembler and then to the original Microsoft Basic, on an Apple II machine with twin floppy discs (5.25 inches and definitely floppy). I used it as part of a system to monitor and control energy use in a drying process for the china clay industry, but the harsh industrial environment very quickly destroyed the floppy discs.
At home I had to moved on to a Dragon 32, made in Wales and based on the Tandy colour computer. This was an interesting machine has it used the Motorola 6809 processor, billed as the first 8 / 16 bit process as it had both 8 bit an d 16 bit registers. This was the first time I cam across multi tasking when Motorola commissioned Microware to develop a multi tasking , multi user operating system to run on the 6809 processor. They produced a version for the Tandy machine and due to the design similarities it also ran on the Dragon, though multi tasking in just 32 kilobytes was a bit limiting. By now I was using the new C programming language, more powerful than BASIC, but with the abilities of assembler code. Microware rewrote the OS9 operating system to work with one of the early 16 bit micro-processors, the 68000 which spawned a host of industrial machines some of which are still in use today. I dumped the old dragon and moved on the the Atari ST with its graphical user interface based on the Apple systems, Apple had settled on the Motorola 68000 series as the basis for their computers and they were streets ahead of the Microsoft DOS based systems of the time.
Unfortunately for us all Microsoft clearly saw into the future and quickly followed Apple to bring out a Windows interface that ran on the DOS operating system culminating in the first workable version - Windows 3.11. As time progressed the war with Apple warmed up and first Windows 95 and then when IBM and Microsoft banded together the early versions of the NT operating system. First as a joint venture producing OS2 and then when Microsoft went its own way they produced NT4 which was the first useable graphical operating system that had the required stability and security for multi user environments. The phenomenal rise of Microsoft soon followed leading to Windows 2000 where they combined the DOS based Windows 98 user interface with the true multi tasking NT technology.
By now the ubiquitous PC was becoming an office standard and many big businesses insisted on using the Windows operating system leading to the demise of many small but more stable operating systems that were much more suited to industrial applications. I was busy implementing various industrial applications in a wide variety of industries rangng from PVC cable manufacture to foundries producing iron castings. These used PLCs from many manufacturers including the GEC GEM80, Allen Bradley PLC 5 and SLC500, Square D, Modicon, Siemens S5 and later S7 and the Telemechanique Series 7. Programming had moved on from dedicated units to PC based software and I rapidly moved through the laptop market firstly with an 12MHz Samsung with a 286 processor (hardly laptop in size) and eventually to current machines based on Intel duo core processors.
OS9 and OS9-68K based computers served me well through the 90's but with the 21st century the PC became all powerful, and operator interfaces moved over to the PC.
I had become heavily involved with a 4GL programming language, called Sculptor, during the early 80s which had the unusual attribute that the binary files were compatible with Xenix, UNIX, DOs and OS9 (OS9 has a Unix like structure but is ram rather than disc based). So I was able to program on the PC but transfer the compiled software to run on OS9 multi tasking units which out performed any DOS based systems. The language is still available today although there is now a Windows flavour, so systems I designed over 20 years ago can still be upgraded if required.
We use the language to design traceability systems as the database is scaleable and extremely reliable over long periods (years). To date we have over sixty applications in the field some dating back to the early 90's